Transition experiments gone wrong